Small JRuby lib using jodconverter to convert any office file to HTML and get it's content.

converter.rb

require 'commons-cli-1.1.jar'
require 'jodconverter-core-3.0-beta-4.jar'
require 'juh-3.2.1.jar'
require 'ridl-3.2.1.jar'
require 'commons-io-1.4.jar'
require 'json-20090211.jar'
require 'jurt-3.2.1.jar'
require 'unoil-3.2.1.jar'

class Converter
  @@manager = false

  def self.start
    @@manager = org.artofsolving.jodconverter.office.DefaultOfficeManagerConfiguration.new.buildOfficeManager
    @@manager.start

    @@manager
  end

  def self.convert(source)
    @@manager ||= self.start

    file = Tempfile.new(['converter', '.html'])
    file.close

    converter = org.artofsolving.jodconverter.OfficeDocumentConverter.new(@@manager)
    converter.convert(java.io.File.new(source), java.io.File.new(file.path))

    file.open
    content = file.read

    file.close
    file.unlink

    return content.force_encoding('BINARY')
  end

  def self.stop
    @@manager ||= self.start

    @@manager.stop
    @@manager = false

    true
  end
end